yamcha
« Back to VersTracker
Description:
NLP text chunker using Support Vector Machines
Type: Formula  |  Latest Version: 0.33@0  |  Tracked Since: Dec 24, 2025
Links: Homepage  |  formulae.brew.sh
Category: Ai ml
Tags: nlp machine-learning japanese svm text-processing
Install: brew install yamcha
About:
Yamcha is a high-performance, general-purpose chunker for natural language processing tasks. It utilizes Support Vector Machines (SVMs) to accurately identify syntactic boundaries, such as morphemes or phrases, within Japanese text. This tool is essential for preparing raw text for downstream NLP applications like parsing or semantic analysis.
Key Features:
  • Support Vector Machine based learning
  • High performance and accuracy
  • Supports Japanese morphological analysis
  • Configurable feature extraction
Use Cases:
  • Pre-processing Japanese text for search engines
  • Training custom morphological analyzers
  • Tokenization for natural language parsing pipelines
Alternatives:
  • MeCab – MeCab is a more widely used Japanese morphological analyzer, while Yamcha specifically focuses on the SVM-based chunking approach.
  • CRF++ – CRF++ uses Conditional Random Fields rather than SVMs, offering a different statistical approach to sequence labeling.
Version History
Detected Version Rev Change Commit
Dec 24, 2025 10:33pm 0.33 0 VERSION_BUMP 3107e9e5
Sep 11, 2025 7:34pm 0 VERSION_BUMP 0827d3a2